The basic concept of the water vapor radiometry experiment at OVRO is conceptually straight forward. The wavefront of an astronomical signal is distorted by fluctuations in the amount of water vapor as it passes through the Earth's atmosphere. These fluctuations are the dominant source of phase noise at millimeter wavelengths, and limit the resolution of millimeter wave arrays at the longest baselines. It seems logical that if one could monitor the time variation in the amount of atmospheric water vapor, one could undo these fluctuations and recover the undistorted astronomical signal. We are attempting to do just that at OVRO by monitoring emission from the atmospheric water line at 22 GHz. We have chosen this particular line (as opposed to the 183 GHz line) since it is optically thin for the observing conditions at Owens Valley.

The OVRO Water Line Monitors (WLM) are designed to isolate emission from the 22 GHz atmospheric water vapor line from sources of continuum emission in the 22 GHz band (e.g. liquid water). This is necessary since for a given 1 Kelvin change in brightness temperature, the phase delay caused by water vapor is about 100 times that due to liquid water. Thus in cloudy weather conditions, liquid droplets in the clouds will modify the total signal in the 22 GHz band, but not contribute any substantial phase noise to the astronomical data.

The OVRO WLM subtract the continuum emission using a three channel system as illustrated in the figure above. The think black lines in this figure are models of the 22 GHz atmospheric water vapor line, and the three vertical rectangles represent the three channels in the WLM system. One channel is centered on the core of the WLM, while the other two channels (in red) are offset +/- 3 GHz from this line and measure the continuum emission. The bandwidth of each channel is 2 GHz. The two continuum channels are averaged and subtracted from the center channel to provide a measure of the emission from the water vapor line itself.